Mmcp.market

credible-claims skill

by pedrohcgs·pedrohcgs/claude-code-my-workflow·1.6k stars·MIT

Research-brief + claim-record discipline for delegated or AI-assisted research work. Use when starting any substantive research task or long autonomous run (write the brief first), and when reporting results that will support a claim in a paper or decision (produce the claim record). Keeps faster execution from being confused with credible evidence.

A100/100content scan

Is the credible-claims skill safe?

Clean: nothing in its files matched our rules. We read 1 file in the folder on 2026-09-28.

No findings.

Install the credible-claims skill

A skill is a folder. Copy it into your agent's skills folder and the agent loads it when the task matches its description. 

git clone --depth 1 https://github.com/pedrohcgs/claude-code-my-workflow.git /tmp/claude-code-my-workflow
mkdir -p ~/.claude/skills
cp -r /tmp/claude-code-my-workflow/.claude/skills/credible-claims ~/.claude/skills/credible-claims
available in every project

In the Claude apps, zip the folder and upload it from the Skills settings. The folder on GitHub

The instructions your agent would load

SKILL.md as published, without the frontmatter. Read it on GitHub

Credible claims: the brief before, the record after

Cheaper generation increases demand for scarce validation. The fix is two lightweight artifacts: a research brief that constrains what the system may do, and a claim record that constrains what may be reported. Applies to agents, workflows, sims, proof patches, data construction — anything delegated.

Three standing rules

  1. Delegate only after inputs, boundaries, and completion criteria are clear. No open-ended "make it better" runs.
  2. Require evidence, not confident conclusions. Every delegated task returns inspectable evidence: locations, diffs, diagnostics, counts, failing cases, logs — never just "done/looks fine."
  3. Escalate anything that changes the economic object, the identifying assumptions, the inferential procedure, or the reporting language. Those decisions return to the researcher (the user), always. A standing user ruling counts as a returned decision; record it.

The research brief (before execution)

One short block, written before launching the work:

  • Question / target: what exactly is being estimated, proved, built.
  • Completion: what counts as done; what results would not answer the question.
  • Prohibited substitutions: what the system may not silently change (estimand, sample, assumptions, statement of a theorem, benchmark spec).
  • Known failure modes: what tends to go wrong here; the checks matched to each.
  • Required evidence: what must come back (numbers, locations, diffs, diagnostics).
  • Escalation triggers: which findings/decisions must return to the user before proceeding.
  • Blocked-route rule: a route that depends on unavailable data, an unsupported assumption, or an unproved result is marked blocked — a scientific outcome, not an instruction to search until a favorable answer appears.

The claim record (during/after execution)

For each claim the work will support:

  • Support: which data/analysis/proof supports it (with locations).
  • Changes after seeing results: anything modified after outcomes were visible, and why (diagnostic-triggered fix vs favorable switch — keep these distinguishable).
  • Unresolved: checks that remain open, and how they constrain the language.
  • Decision: who decided what could be reported (user ruling vs assistant default).

Proportionality: a routine task needs one short paragraph; heavier records only when branching is extensive, outputs will be reused, errors are consequential, or correction is costly.

Reporting language

The final decision is never "all checks green" — it is whether the evidence supports the proposed language. The options are: repair; narrower language; additional review; an exploratory/descriptive label; or decline to report. Never upgrade language beyond the evidence (associational ≠ causal; pointwise ≠ uniform; illustrated ≠ validated; imposed ≠ derived).

Keep credibility questions separate

Reproducibility, implementation correctness, statistical performance, measurement validity, and identification/scope are different questions; evidence on one cannot answer another. Reproducible code may implement the wrong estimator; favorable simulations cannot establish an assumption; a correct estimate may answer the wrong question.

Learn forward

Every diagnosed failure becomes a durable artifact: a reusable test, a documented warning, or a memory entry stating the failure, why it happened, and the check that now prevents it. Preserve failed approaches and the reason they failed — a blocked route re-attempted without a new mechanism is waste.

Cross-references

  • verification-ladder.md — rung 5 (the ledger)
  • .claude/rules/orchestrator-protocol.md — RUN_CONFIG is the brief for a fan-out

More skills from pedrohcgs/claude-code-my-workflow

  • Aadjudicate-reviewTurn an incoming set of findings — from an AI reviewer, a referee report, a code review, a linter, or a second model — into verified fixes, without letting a confident misread damage correct work. Every finding is a CANDIDATE until checked against the actual source. Use whenever you receive review comments, audit findings, or a critique you did not write yourself, especially when the reviewer is a model or when the volume is too large to check by feel.
  • Aaudit-reproducibilityEnforce the replication-protocol.md rule by cross-checking numeric claims in a manuscript against the actual R / Stata / Python outputs. Report PASS/FAIL per claim against tolerance thresholds. Use before submission and before releasing a replication package.
  • Ablast-radiusBefore and after changing anything shared — a function's return value, a signature, a schema, a label set, a config default, a constant, a file format — find every consumer and actually run them. Catches the change that looks purely additive but silently breaks a contract in a file you never opened. Use when editing shared code, adding a field/column/return element, renaming, changing units or defaults, or touching a pipeline that produces reported numbers.
  • Acapture-environmentSnapshot the computational environment for a replication package — detects the analysis stack (R / Stata / Python) and emits the right lockfiles (renv.lock + sessionInfo.txt, requirements.txt / environment.yml / uv.lock, Stata version + ado package list), records seeds and RNG kind, optionally writes a pinning Dockerfile, and produces a paste-ready "Computational requirements" block. Use when user says "capture the environment", "snapshot my dependencies", "pin the versions", "make a renv.lock / requirements.txt", "make this byte-reproducible", or before releasing a replication package to openICPSR / the AEA Data Editor.
  • AchallengeStress-test a finding against the choices you did not make. Enumerates the discrete forks a competent analyst could have taken (measure definition, sample filter, control set, clustering level, weighting, functional form), runs the specification grid, and reports the distribution rather than a point estimate — then attacks the identifying assumption with named, computable sensitivity statistics. Use when the user says "is this robust", "challenge this result", "specification curve", "multiverse", "how sensitive is this", "what if I'd used a different measure", "stress-test my estimate", or before a result becomes a headline claim. NOT a reviewer of prose or code — it challenges the CLAIM.
  • AcheckpointSave a structured state snapshot before stopping or handing off. Captures the active plan, recent decisions, file pointers (with line numbers), open questions, and the next 1–3 actions into a checkpoint file under `quality_reports/checkpoints/`. Optionally proposes `[LEARN]` entries to add to MEMORY.md. Use when user says "checkpoint", "save state", "snapshot before I stop", "where am I", "wrap up the session for handoff", or before a long break / model switch / collaborator handoff. Companion to (NOT replacement for) the narrative session-log workflow.
  • Acoauthor-briefGenerate a co-author / collaborator handoff brief for a multi-author, multi-machine project — summarizing what changed since the last brief (git delta), the current state of each artifact (manuscript, analysis, slides), open questions, how to reproduce locally, and any restricted-data access steps. Use when user says "coauthor brief", "handoff brief", "bring my coauthor up to speed", "what changed since last week", "onboard a collaborator", "write a handoff for [name]", or before sending a co-author the repo. NOT a commit or a checkpoint — it is the cross-machine, cross-person summary `meta-governance.md` only partially covers.
  • AcommitCommit the current work — runs the quality, consistency and passport gates, branches off main if needed, stages specific files, and writes a commit whose subject states what is now true. Pushes and opens a pull request only with --pr or when the user asks; never merges — a merge happens only when the user explicitly says to merge. Use ONLY on explicit commit intent — user says "commit", "let's commit this", "open a PR", or prefixes with `/commit`. Do NOT auto-invoke on vague end-of-task phrases ("we're done", "wrap up") — those require explicit confirmation first. Never force-pushes or skips hooks.
  • Acompile-latexCompile a Beamer LaTeX slide deck with XeLaTeX (3 passes + bibtex). Use when user says "compile", "build the slides", "rebuild the PDF", "run latex", "render the tex", or asks why a `.tex` file isn't producing a PDF. Operates on `Slides/*.tex`.
  • Acompress-sessionDistill the current conversation into a structured note (decisions made, open questions, file pointers with line numbers, next 1–3 actions) and save to `quality_reports/session_logs/` before auto-compression. Differs from `/checkpoint` (explicit stop-point snapshot) and from auto-compaction (which truncates rather than distills). Use when context is approaching auto-compact threshold, when a long pipeline has accumulated many decisions, or when the user says "compress", "distil this session", "before we hit auto-compact", "structured handoff before context resets".
  • Acontext-statusShow current context status and session health. Use to check how much context has been used, whether auto-compact is approaching, and what state will be preserved.
  • Acreate-lectureCreate a new Beamer lecture `.tex` from source papers and materials, with notation consistency checks and the project's preamble wired in. Use when user says "create a lecture on X", "new lecture from these papers", "start a deck on topic Y", "scaffold a new Beamer file", "build me a lecture from these PDFs". Scaffolds the full deck — NOT for compiling existing `.tex` (use `/compile-latex`).

All agent skills → · MCP servers